feat(hashql-core): add LinkedGraph benchmarking suite with instruction counting by indietyp · Pull Request #8420 · hashintel/hash

indietyp · 2026-02-13T16:36:57Z

Summary

Add a comprehensive benchmarking suite for the LinkedGraph data structure in hashql-core. The benchmarks use instruction counting on Linux (via perf_event) with a fallback to wall-clock time on other platforms.

What's Included

Benchmark Groups

node: add_node, lookup by ID, iteration over nodes
edge: add_edge, lookup by ID, iteration over edges
adjacency: successors, predecessors, incident edges iteration
traversal: DFS, BFS, post-order traversal across various graph topologies
mutation: clear_edges, clear, derive operations

Graph Fixtures

Chain graph: Linear 0 → 1 → 2 → ... → n-1
Complete graph: Every node connected to every other
Binary tree: Balanced tree of configurable depth
Sparse graph: Fixed edges per node with deterministic distribution

Implementation Details

Instruction Counting (Linux)

Uses the perf_event crate to create a hardware instruction counter. This provides deterministic, reproducible benchmark results that are independent of system load.

let counter = Builder::new()
    .kind(Hardware::INSTRUCTIONS)
    .build()?;

Fallback (Non-Linux)

On platforms that don't support perf_event (macOS, Windows), the benchmarks automatically fall back to criterion's default wall-clock time measurement.

How to Run

# Run all linked graph benchmarks
cargo bench --package hashql-core --bench linked_graph

# Run specific benchmark group
cargo bench --package hashql-core --bench linked_graph -- "node"
cargo bench --package hashql-core --bench linked_graph -- "traversal"

Example Output (Linux with instruction counting)

node/add/10000          time:   [197.12 Ki 197.14 Ki 197.15 Ki]
                        thrpt:  [19.715 inst/elem 19.714 inst/elem 19.712 inst/elem]

node/iter/1000          time:   [4.0700 Ki 4.0700 Ki 4.0700 Ki]
                        thrpt:  [4.0700 inst/elem 4.0700 inst/elem 4.0700 inst/elem]

This PR was generated with Warp.

…n counting Add a comprehensive benchmarking suite for the LinkedGraph data structure that uses instruction counting on Linux (via perf_event) with a fallback to wall-clock time on other platforms. Benchmark groups: - node: add, lookup, iteration operations - edge: add, lookup, iteration operations - adjacency: successors, predecessors, incident edges iteration - traversal: DFS, BFS, post-order traversal on various topologies - mutation: clear_edges, clear, derive operations Graph fixtures include: chain, complete, binary tree, and sparse graphs. Co-Authored-By: Warp <agent@warp.dev>

cursor · 2026-02-13T16:37:04Z

PR Summary

Low Risk
Changes are limited to dev-dependencies, lockfile updates, and new benchmark code; no production logic is modified, with the main risk being CI/benchmark build issues on Linux due to perf-event permissions or platform differences.

Overview
Adds a new hashql-core benchmark suite (benches/linked_graph.rs) covering LinkedGraph node/edge operations, adjacency iteration, traversals, and mutation APIs across several graph fixtures.

On Linux, benchmarks now optionally measure CPU instruction counts via perf-event (custom Criterion Measurement + formatter), with a wall-clock fallback on other platforms; workspace/dev-dependencies are updated accordingly and linked_graph is registered as a new cargo bench target.

^{Written by Cursor Bugbot for commit 1c41c5f. This will update automatically on new commits. Configure here.}

vercel · 2026-02-13T16:37:04Z

The latest updates on your projects. Learn more about Vercel for GitHub.

Project	Deployment	Actions	Updated (UTC)
hash	Ready	Preview, Comment	Feb 13, 2026 4:49pm
hashdotdesign	Ready	Preview, Comment	Feb 13, 2026 4:49pm
hashdotdesign-tokens	Ready	Preview, Comment	Feb 13, 2026 4:49pm
petrinaut	Ready	Preview, Comment	Feb 13, 2026 4:49pm

augmentcode · 2026-02-13T16:39:00Z

🤖 Augment PR Summary

Summary: Adds a new Criterion benchmarking suite for the hashql-core LinkedGraph data structure, with Linux instruction-count measurement.

Changes:

Adds perf-event to workspace dependencies and wires it as a Linux-only dev-dependency for hashql-core
Registers a new linked_graph bench target (non-harness) alongside the existing benches
Introduces benches/linked_graph.rs covering node, edge, adjacency, traversal, and mutation benchmark groups
Implements a Linux-only Criterion Measurement that counts CPU instructions via perf_event
Uses wall-clock time (WallTime) automatically on non-Linux targets
Adds deterministic graph fixtures (chain, complete, binary tree, sparse) to exercise different topologies

Technical Notes: On Linux, measurements are taken from a hardware instruction counter (reset/enable per iteration, disable/read at end). Throughput is primarily reported as instructions per element using Criterion’s throughput reporting.

_{🤖 Was this summary useful? React with 👍 or 👎}

augmentcode

Review completed. 2 suggestions posted.

Comment augment review to trigger a new review at any time.

augmentcode · 2026-02-13T16:39:01Z

libs/@local/hashql/core/benches/linked_graph.rs

+            let counter = Builder::new()
+                .kind(Hardware::INSTRUCTIONS)
+                .build()
+                .expect(


InstructionCount::new() currently expects on perf_event counter creation, which will make cargo bench fail on many Linux setups (e.g., containers/CI with restrictive perf_event_paranoid). Consider a graceful fallback (or an opt-in flag) so benchmarks still run when hardware counters aren’t available.

Severity: medium

_{🤖 Was this useful? React with 👍 or 👎, or 🚀 if it prevented an incident/outage.}

augmentcode · 2026-02-13T16:39:02Z

libs/@local/hashql/core/benches/linked_graph.rs

+                criterion::Throughput::Bytes(bytes)
+                | criterion::Throughput::BytesDecimal(bytes) => {
+                    for value in values {
+                        *value = (*bytes as f64) / *value;


In scale_throughputs for Throughput::Bytes*, the calculation is inverted: *value = bytes / value yields bytes-per-instruction, but the unit returned is inst/B. If you intend inst/B, this should scale like the Elements case (divide instructions by bytes).

Severity: low

_{🤖 Was this useful? React with 👍 or 👎, or 🚀 if it prevented an incident/outage.}

github-advanced-security

clippy found more than 20 potential problems in the proposed changes. Check the Files changed tab for more details.

codecov · 2026-02-13T16:40:31Z

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 60.99%. Comparing base (c8e4ac9) to head (1c41c5f).
⚠️ Report is 8 commits behind head on main.

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #8420      +/-   ##
==========================================
- Coverage   61.05%   60.99%   -0.07%     
==========================================
  Files        1247     1248       +1     
  Lines      121606   121739     +133     
  Branches     5267     5299      +32     
==========================================
+ Hits        74245    74251       +6     
- Misses      46489    46616     +127     
  Partials      872      872

Flag	Coverage Δ
apps.hash-ai-worker-ts	`1.40% <ø> (-0.01%)`	⬇️
apps.hash-api	`0.00% <ø> (ø)`
blockprotocol.type-system	`40.84% <ø> (ø)`
local.claude-hooks	`0.00% <ø> (ø)`
local.harpc-client	`51.24% <ø> (ø)`
local.hash-graph-sdk	`7.78% <ø> (+1.11%)`	⬆️
local.hash-isomorphic-utils	`0.00% <ø> (ø)`
rust.antsi	`0.00% <ø> (ø)`
rust.error-stack	`90.88% <ø> (ø)`
rust.harpc-codec	`84.70% <ø> (ø)`
rust.harpc-net	`96.18% <ø> (+0.03%)`	⬆️
rust.harpc-tower	`66.80% <ø> (ø)`
rust.harpc-types	`0.00% <ø> (ø)`
rust.harpc-wire-protocol	`92.23% <ø> (ø)`
rust.hash-codec	`72.76% <ø> (ø)`
rust.hash-graph-api	`2.88% <ø> (ø)`
rust.hash-graph-authorization	`62.34% <ø> (ø)`
rust.hash-graph-postgres-store	`26.80% <ø> (ø)`
rust.hash-graph-store	`39.10% <ø> (ø)`
rust.hash-graph-temporal-versioning	`47.95% <ø> (ø)`
rust.hash-graph-types	`0.00% <ø> (ø)`
rust.hash-graph-validation	`83.45% <ø> (ø)`
rust.hashql-ast	`87.25% <ø> (ø)`
rust.hashql-compiletest	`46.65% <ø> (ø)`
rust.hashql-core	`81.75% <ø> (ø)`
rust.hashql-diagnostics	`72.43% <ø> (ø)`
rust.hashql-eval	`68.54% <ø> (ø)`
rust.hashql-hir	`89.10% <ø> (ø)`
rust.hashql-mir	`89.51% <ø> (ø)`
rust.hashql-syntax-jexpr	`94.05% <ø> (ø)`

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.
📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.

codspeed-hq · 2026-02-13T16:42:53Z

Merging this PR will not alter performance

✅ 21 untouched benchmarks

_{Comparing feat/linked-graph-benchmarks (1c41c5f) with main (c8e4ac9)¹}

No successful run was found on main (c5edd55) during the generation of this report, so c8e4ac9 was used instead as the comparison base. There might be some changes unrelated to this pull request in this report. ↩

cursor

Cursor Bugbot has reviewed your changes and found 2 potential issues.

^{Bugbot Autofix is OFF. To automatically fix reported issues with Cloud Agents, enable Autofix in the Cursor dashboard.}

cursor · 2026-02-13T16:43:55Z

libs/@local/hashql/core/benches/linked_graph.rs

+                        *value = (*bytes as f64) / *value;
+                    }
+                    "inst/B"
+                }


Inverted throughput formula for Bytes measurement scaling

Medium Severity

The Bytes/BytesDecimal branch in scale_throughputs computes bytes / instructions (bytes per instruction) but labels the result "inst/B" (instructions per byte). The formula is the reciprocal of what the unit label indicates. The Elements branch correctly uses *value /= *elements as f64 to get instructions-per-element, but the Bytes branch inconsistently uses *value = (*bytes as f64) / *value instead of the analogous *value /= *bytes as f64.

cursor · 2026-02-13T16:43:55Z

libs/@local/hashql/core/benches/linked_graph.rs

+            let mid = NodeId::new(size / 2);
+            b.iter(|| black_box(graph.node(mid)));
+        });
+    }


Stale throughput leaks into lookup benchmark results

Low Severity

The group.throughput() call from the preceding "add" loop persists on the BenchmarkGroup and leaks into the "lookup" benchmarks, which don't reset it. In bench_node_operations, lookup benchmarks inherit a stale throughput of 10,000 elements; in bench_edge_operations, they inherit 249,500 elements. Since lookup measures a single O(1) operation, the reported throughput (inst/elem) values will be orders of magnitude off, producing misleading benchmark output.

Additional Locations (1)

libs/@local/hashql/core/benches/linked_graph.rs#L387-L395

github-actions · 2026-02-13T17:50:08Z

Benchmark results

@rust/hash-graph-benches – Integrations

policy_resolution_large

Function	Value	Mean	Flame graphs
resolve_policies_for_actor	user: empty, selectivity: high, policies: 2002	$$27.2 \mathrm{ms} \pm 156 \mathrm{μs}\left({\color{gray}0.016 \mathrm{\%}}\right) $$	Flame Graph
resolve_policies_for_actor	user: empty, selectivity: low, policies: 1	$$3.48 \mathrm{ms} \pm 16.9 \mathrm{μs}\left({\color{gray}-0.011 \mathrm{\%}}\right) $$	Flame Graph
resolve_policies_for_actor	user: empty, selectivity: medium, policies: 1001	$$12.3 \mathrm{ms} \pm 76.2 \mathrm{μs}\left({\color{gray}-0.406 \mathrm{\%}}\right) $$	Flame Graph
resolve_policies_for_actor	user: seeded, selectivity: high, policies: 3314	$$56.7 \mathrm{ms} \pm 423 \mathrm{μs}\left({\color{red}32.1 \mathrm{\%}}\right) $$	Flame Graph
resolve_policies_for_actor	user: seeded, selectivity: low, policies: 1	$$14.7 \mathrm{ms} \pm 88.8 \mathrm{μs}\left({\color{gray}-0.424 \mathrm{\%}}\right) $$	Flame Graph
resolve_policies_for_actor	user: seeded, selectivity: medium, policies: 1526	$$24.3 \mathrm{ms} \pm 165 \mathrm{μs}\left({\color{gray}0.868 \mathrm{\%}}\right) $$	Flame Graph
resolve_policies_for_actor	user: system, selectivity: high, policies: 2078	$$28.4 \mathrm{ms} \pm 238 \mathrm{μs}\left({\color{gray}1.42 \mathrm{\%}}\right) $$	Flame Graph
resolve_policies_for_actor	user: system, selectivity: low, policies: 1	$$3.78 \mathrm{ms} \pm 15.6 \mathrm{μs}\left({\color{gray}-0.312 \mathrm{\%}}\right) $$	Flame Graph
resolve_policies_for_actor	user: system, selectivity: medium, policies: 1033	$$13.2 \mathrm{ms} \pm 71.5 \mathrm{μs}\left({\color{gray}-0.558 \mathrm{\%}}\right) $$	Flame Graph

policy_resolution_medium

Function	Value	Mean	Flame graphs
resolve_policies_for_actor	user: empty, selectivity: high, policies: 102	$$3.90 \mathrm{ms} \pm 22.8 \mathrm{μs}\left({\color{gray}0.125 \mathrm{\%}}\right) $$	Flame Graph
resolve_policies_for_actor	user: empty, selectivity: low, policies: 1	$$3.09 \mathrm{ms} \pm 16.3 \mathrm{μs}\left({\color{gray}0.496 \mathrm{\%}}\right) $$	Flame Graph
resolve_policies_for_actor	user: empty, selectivity: medium, policies: 51	$$3.42 \mathrm{ms} \pm 16.6 \mathrm{μs}\left({\color{gray}-0.609 \mathrm{\%}}\right) $$	Flame Graph
resolve_policies_for_actor	user: seeded, selectivity: high, policies: 269	$$5.30 \mathrm{ms} \pm 26.2 \mathrm{μs}\left({\color{gray}-0.727 \mathrm{\%}}\right) $$	Flame Graph
resolve_policies_for_actor	user: seeded, selectivity: low, policies: 1	$$3.70 \mathrm{ms} \pm 26.4 \mathrm{μs}\left({\color{gray}-0.016 \mathrm{\%}}\right) $$	Flame Graph
resolve_policies_for_actor	user: seeded, selectivity: medium, policies: 107	$$4.25 \mathrm{ms} \pm 21.7 \mathrm{μs}\left({\color{gray}0.058 \mathrm{\%}}\right) $$	Flame Graph
resolve_policies_for_actor	user: system, selectivity: high, policies: 133	$$4.58 \mathrm{ms} \pm 27.4 \mathrm{μs}\left({\color{gray}0.961 \mathrm{\%}}\right) $$	Flame Graph
resolve_policies_for_actor	user: system, selectivity: low, policies: 1	$$3.52 \mathrm{ms} \pm 18.7 \mathrm{μs}\left({\color{gray}0.671 \mathrm{\%}}\right) $$	Flame Graph
resolve_policies_for_actor	user: system, selectivity: medium, policies: 63	$$4.21 \mathrm{ms} \pm 33.9 \mathrm{μs}\left({\color{gray}1.33 \mathrm{\%}}\right) $$	Flame Graph

policy_resolution_none

Function	Value	Mean	Flame graphs
resolve_policies_for_actor	user: empty, selectivity: high, policies: 2	$$2.80 \mathrm{ms} \pm 16.2 \mathrm{μs}\left({\color{gray}-0.325 \mathrm{\%}}\right) $$	Flame Graph
resolve_policies_for_actor	user: empty, selectivity: low, policies: 1	$$2.78 \mathrm{ms} \pm 15.3 \mathrm{μs}\left({\color{gray}1.37 \mathrm{\%}}\right) $$	Flame Graph
resolve_policies_for_actor	user: empty, selectivity: medium, policies: 1	$$2.87 \mathrm{ms} \pm 15.1 \mathrm{μs}\left({\color{gray}-0.099 \mathrm{\%}}\right) $$	Flame Graph
resolve_policies_for_actor	user: system, selectivity: high, policies: 8	$$3.15 \mathrm{ms} \pm 21.7 \mathrm{μs}\left({\color{gray}0.995 \mathrm{\%}}\right) $$	Flame Graph
resolve_policies_for_actor	user: system, selectivity: low, policies: 1	$$2.94 \mathrm{ms} \pm 16.4 \mathrm{μs}\left({\color{gray}-0.141 \mathrm{\%}}\right) $$	Flame Graph
resolve_policies_for_actor	user: system, selectivity: medium, policies: 3	$$3.22 \mathrm{ms} \pm 21.5 \mathrm{μs}\left({\color{gray}-0.481 \mathrm{\%}}\right) $$	Flame Graph

policy_resolution_small

Function	Value	Mean	Flame graphs
resolve_policies_for_actor	user: empty, selectivity: high, policies: 52	$$3.15 \mathrm{ms} \pm 15.8 \mathrm{μs}\left({\color{gray}-0.342 \mathrm{\%}}\right) $$	Flame Graph
resolve_policies_for_actor	user: empty, selectivity: low, policies: 1	$$2.88 \mathrm{ms} \pm 14.2 \mathrm{μs}\left({\color{gray}-0.094 \mathrm{\%}}\right) $$	Flame Graph
resolve_policies_for_actor	user: empty, selectivity: medium, policies: 25	$$3.05 \mathrm{ms} \pm 14.9 \mathrm{μs}\left({\color{gray}0.238 \mathrm{\%}}\right) $$	Flame Graph
resolve_policies_for_actor	user: seeded, selectivity: high, policies: 94	$$3.63 \mathrm{ms} \pm 20.8 \mathrm{μs}\left({\color{gray}-0.069 \mathrm{\%}}\right) $$	Flame Graph
resolve_policies_for_actor	user: seeded, selectivity: low, policies: 1	$$3.14 \mathrm{ms} \pm 13.6 \mathrm{μs}\left({\color{gray}-0.015 \mathrm{\%}}\right) $$	Flame Graph
resolve_policies_for_actor	user: seeded, selectivity: medium, policies: 26	$$3.39 \mathrm{ms} \pm 14.5 \mathrm{μs}\left({\color{gray}-1.438 \mathrm{\%}}\right) $$	Flame Graph
resolve_policies_for_actor	user: system, selectivity: high, policies: 66	$$3.52 \mathrm{ms} \pm 16.2 \mathrm{μs}\left({\color{gray}-0.382 \mathrm{\%}}\right) $$	Flame Graph
resolve_policies_for_actor	user: system, selectivity: low, policies: 1	$$3.13 \mathrm{ms} \pm 16.8 \mathrm{μs}\left({\color{gray}-1.002 \mathrm{\%}}\right) $$	Flame Graph
resolve_policies_for_actor	user: system, selectivity: medium, policies: 29	$$3.42 \mathrm{ms} \pm 18.8 \mathrm{μs}\left({\color{gray}-1.827 \mathrm{\%}}\right) $$	Flame Graph

read_scaling_complete

Function	Value	Mean	Flame graphs
entity_by_id;one_depth	1 entities	$$41.1 \mathrm{ms} \pm 188 \mathrm{μs}\left({\color{gray}0.112 \mathrm{\%}}\right) $$	Flame Graph
entity_by_id;one_depth	10 entities	$$78.9 \mathrm{ms} \pm 387 \mathrm{μs}\left({\color{gray}1.19 \mathrm{\%}}\right) $$	Flame Graph
entity_by_id;one_depth	25 entities	$$46.4 \mathrm{ms} \pm 202 \mathrm{μs}\left({\color{gray}2.91 \mathrm{\%}}\right) $$	Flame Graph
entity_by_id;one_depth	5 entities	$$48.1 \mathrm{ms} \pm 232 \mathrm{μs}\left({\color{gray}0.844 \mathrm{\%}}\right) $$	Flame Graph
entity_by_id;one_depth	50 entities	$$55.4 \mathrm{ms} \pm 355 \mathrm{μs}\left({\color{gray}-0.148 \mathrm{\%}}\right) $$	Flame Graph
entity_by_id;two_depth	1 entities	$$43.1 \mathrm{ms} \pm 195 \mathrm{μs}\left({\color{gray}1.58 \mathrm{\%}}\right) $$	Flame Graph
entity_by_id;two_depth	10 entities	$$426 \mathrm{ms} \pm 877 \mathrm{μs}\left({\color{gray}2.18 \mathrm{\%}}\right) $$	Flame Graph
entity_by_id;two_depth	25 entities	$$92.3 \mathrm{ms} \pm 537 \mathrm{μs}\left({\color{lightgreen}-5.372 \mathrm{\%}}\right) $$	Flame Graph
entity_by_id;two_depth	5 entities	$$88.2 \mathrm{ms} \pm 411 \mathrm{μs}\left({\color{gray}2.31 \mathrm{\%}}\right) $$	Flame Graph
entity_by_id;two_depth	50 entities	$$283 \mathrm{ms} \pm 933 \mathrm{μs}\left({\color{lightgreen}-9.823 \mathrm{\%}}\right) $$	Flame Graph
entity_by_id;zero_depth	1 entities	$$15.8 \mathrm{ms} \pm 73.7 \mathrm{μs}\left({\color{gray}0.788 \mathrm{\%}}\right) $$	Flame Graph
entity_by_id;zero_depth	10 entities	$$15.7 \mathrm{ms} \pm 88.6 \mathrm{μs}\left({\color{gray}-0.878 \mathrm{\%}}\right) $$	Flame Graph
entity_by_id;zero_depth	25 entities	$$15.9 \mathrm{ms} \pm 83.3 \mathrm{μs}\left({\color{gray}-0.289 \mathrm{\%}}\right) $$	Flame Graph
entity_by_id;zero_depth	5 entities	$$15.9 \mathrm{ms} \pm 99.1 \mathrm{μs}\left({\color{gray}2.57 \mathrm{\%}}\right) $$	Flame Graph
entity_by_id;zero_depth	50 entities	$$18.0 \mathrm{ms} \pm 128 \mathrm{μs}\left({\color{gray}-4.051 \mathrm{\%}}\right) $$	Flame Graph

read_scaling_linkless

Function	Value	Mean	Flame graphs
entity_by_id	1 entities	$$15.3 \mathrm{ms} \pm 60.5 \mathrm{μs}\left({\color{gray}-2.667 \mathrm{\%}}\right) $$	Flame Graph
entity_by_id	10 entities	$$15.6 \mathrm{ms} \pm 90.9 \mathrm{μs}\left({\color{gray}0.408 \mathrm{\%}}\right) $$	Flame Graph
entity_by_id	100 entities	$$15.6 \mathrm{ms} \pm 69.8 \mathrm{μs}\left({\color{gray}0.046 \mathrm{\%}}\right) $$	Flame Graph
entity_by_id	1000 entities	$$16.2 \mathrm{ms} \pm 94.3 \mathrm{μs}\left({\color{gray}0.470 \mathrm{\%}}\right) $$	Flame Graph
entity_by_id	10000 entities	$$23.5 \mathrm{ms} \pm 158 \mathrm{μs}\left({\color{gray}-1.109 \mathrm{\%}}\right) $$	Flame Graph

representative_read_entity

Function	Value	Mean	Flame graphs
entity_by_id	entity type ID: `https://blockprotocol.org/@alice/types/entity-type/block/v/1`	$$30.4 \mathrm{ms} \pm 246 \mathrm{μs}\left({\color{gray}-2.677 \mathrm{\%}}\right) $$	Flame Graph
entity_by_id	entity type ID: `https://blockprotocol.org/@alice/types/entity-type/book/v/1`	$$31.0 \mathrm{ms} \pm 284 \mathrm{μs}\left({\color{gray}-2.597 \mathrm{\%}}\right) $$	Flame Graph
entity_by_id	entity type ID: `https://blockprotocol.org/@alice/types/entity-type/building/v/1`	$$30.5 \mathrm{ms} \pm 309 \mathrm{μs}\left({\color{gray}-0.751 \mathrm{\%}}\right) $$	Flame Graph
entity_by_id	entity type ID: `https://blockprotocol.org/@alice/types/entity-type/organization/v/1`	$$29.8 \mathrm{ms} \pm 252 \mathrm{μs}\left({\color{gray}-3.203 \mathrm{\%}}\right) $$	Flame Graph
entity_by_id	entity type ID: `https://blockprotocol.org/@alice/types/entity-type/page/v/2`	$$31.6 \mathrm{ms} \pm 281 \mathrm{μs}\left({\color{gray}4.75 \mathrm{\%}}\right) $$	Flame Graph
entity_by_id	entity type ID: `https://blockprotocol.org/@alice/types/entity-type/person/v/1`	$$31.2 \mathrm{ms} \pm 281 \mathrm{μs}\left({\color{gray}0.967 \mathrm{\%}}\right) $$	Flame Graph
entity_by_id	entity type ID: `https://blockprotocol.org/@alice/types/entity-type/playlist/v/1`	$$31.0 \mathrm{ms} \pm 277 \mathrm{μs}\left({\color{gray}3.30 \mathrm{\%}}\right) $$	Flame Graph
entity_by_id	entity type ID: `https://blockprotocol.org/@alice/types/entity-type/song/v/1`	$$31.0 \mathrm{ms} \pm 315 \mathrm{μs}\left({\color{gray}-0.735 \mathrm{\%}}\right) $$	Flame Graph
entity_by_id	entity type ID: `https://blockprotocol.org/@alice/types/entity-type/uk-address/v/1`	$$31.0 \mathrm{ms} \pm 335 \mathrm{μs}\left({\color{gray}-2.945 \mathrm{\%}}\right) $$	Flame Graph

representative_read_entity_type

Function	Value	Mean	Flame graphs
get_entity_type_by_id	Account ID: `bf5a9ef5-dc3b-43cf-a291-6210c0321eba`	$$8.60 \mathrm{ms} \pm 41.1 \mathrm{μs}\left({\color{gray}0.177 \mathrm{\%}}\right) $$	Flame Graph

representative_read_multiple_entities

Function	Value	Mean	Flame graphs
entity_by_property	traversal_paths=0	0	$$88.9 \mathrm{ms} \pm 438 \mathrm{μs}\left({\color{gray}-0.014 \mathrm{\%}}\right) $$
entity_by_property	traversal_paths=255	1,resolve_depths=inherit:1;values:255;properties:255;links:127;link_dests:126;type:true	$$141 \mathrm{ms} \pm 530 \mathrm{μs}\left({\color{gray}-0.419 \mathrm{\%}}\right) $$
entity_by_property	traversal_paths=2	1,resolve_depths=inherit:0;values:0;properties:0;links:0;link_dests:0;type:false	$$95.3 \mathrm{ms} \pm 435 \mathrm{μs}\left({\color{gray}-0.940 \mathrm{\%}}\right) $$
entity_by_property	traversal_paths=2	1,resolve_depths=inherit:0;values:0;properties:0;links:1;link_dests:0;type:true	$$104 \mathrm{ms} \pm 479 \mathrm{μs}\left({\color{gray}-0.327 \mathrm{\%}}\right) $$
entity_by_property	traversal_paths=2	1,resolve_depths=inherit:0;values:0;properties:2;links:1;link_dests:0;type:true	$$113 \mathrm{ms} \pm 504 \mathrm{μs}\left({\color{gray}-0.327 \mathrm{\%}}\right) $$
entity_by_property	traversal_paths=2	1,resolve_depths=inherit:0;values:2;properties:2;links:1;link_dests:0;type:true	$$128 \mathrm{ms} \pm 3.30 \mathrm{ms}\left({\color{red}6.00 \mathrm{\%}}\right) $$
link_by_source_by_property	traversal_paths=0	0	$$86.9 \mathrm{ms} \pm 354 \mathrm{μs}\left({\color{gray}-1.007 \mathrm{\%}}\right) $$
link_by_source_by_property	traversal_paths=255	1,resolve_depths=inherit:1;values:255;properties:255;links:127;link_dests:126;type:true	$$114 \mathrm{ms} \pm 504 \mathrm{μs}\left({\color{gray}-1.308 \mathrm{\%}}\right) $$
link_by_source_by_property	traversal_paths=2	1,resolve_depths=inherit:0;values:0;properties:0;links:0;link_dests:0;type:false	$$93.3 \mathrm{ms} \pm 370 \mathrm{μs}\left({\color{gray}-1.142 \mathrm{\%}}\right) $$
link_by_source_by_property	traversal_paths=2	1,resolve_depths=inherit:0;values:0;properties:0;links:1;link_dests:0;type:true	$$106 \mathrm{ms} \pm 1.70 \mathrm{ms}\left({\color{gray}3.14 \mathrm{\%}}\right) $$
link_by_source_by_property	traversal_paths=2	1,resolve_depths=inherit:0;values:0;properties:2;links:1;link_dests:0;type:true	$$103 \mathrm{ms} \pm 450 \mathrm{μs}\left({\color{gray}-1.329 \mathrm{\%}}\right) $$
link_by_source_by_property	traversal_paths=2	1,resolve_depths=inherit:0;values:2;properties:2;links:1;link_dests:0;type:true	$$103 \mathrm{ms} \pm 423 \mathrm{μs}\left({\color{gray}-1.796 \mathrm{\%}}\right) $$

scenarios

Function	Value	Mean	Flame graphs
full_test	query-limited	$$134 \mathrm{ms} \pm 532 \mathrm{μs}\left({\color{gray}1.15 \mathrm{\%}}\right) $$	Flame Graph
full_test	query-unlimited	$$135 \mathrm{ms} \pm 600 \mathrm{μs}\left({\color{gray}3.25 \mathrm{\%}}\right) $$	Flame Graph
linked_queries	query-limited	$$104 \mathrm{ms} \pm 548 \mathrm{μs}\left({\color{gray}0.263 \mathrm{\%}}\right) $$	Flame Graph
linked_queries	query-unlimited	$$590 \mathrm{ms} \pm 3.58 \mathrm{ms}\left({\color{gray}3.51 \mathrm{\%}}\right) $$	Flame Graph

indietyp · 2026-02-13T18:04:53Z

Warp, creating a PR in my name, without asking me is a bad bad thing.

graphite-app bot assigned indietyp Feb 13, 2026

github-actions bot added area/deps Relates to third-party dependencies (area) area/libs Relates to first-party libraries/crates/packages (area) type/eng > backend Owned by the @backend team labels Feb 13, 2026

augmentcode bot reviewed Feb 13, 2026

View reviewed changes

vercel bot deployed to Preview – hashdotdesign-tokens February 13, 2026 16:39 View deployment

github-advanced-security bot found potential problems Feb 13, 2026

View reviewed changes

vercel bot deployed to Preview – hashdotdesign February 13, 2026 16:40 View deployment

vercel bot deployed to Preview – petrinaut February 13, 2026 16:41 View deployment

cursor bot reviewed Feb 13, 2026

View reviewed changes

vercel bot deployed to Preview – hash February 13, 2026 16:49 View deployment

indietyp closed this Feb 13, 2026

Conversation

indietyp commented Feb 13, 2026

Summary

What's Included

Benchmark Groups

Graph Fixtures

Implementation Details

Instruction Counting (Linux)

Fallback (Non-Linux)

How to Run

Example Output (Linux with instruction counting)

Uh oh!

cursor bot commented Feb 13, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

PR Summary

Uh oh!

vercel bot commented Feb 13, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

augmentcode bot commented Feb 13, 2026

Uh oh!

augmentcode bot left a comment

Choose a reason for hiding this comment

Uh oh!

augmentcode bot Feb 13, 2026

Choose a reason for hiding this comment

Uh oh!

augmentcode bot Feb 13, 2026

Choose a reason for hiding this comment

Uh oh!

github-advanced-security bot left a comment

Choose a reason for hiding this comment

Uh oh!

codecov bot commented Feb 13, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Codecov Report

Uh oh!

codspeed-hq bot commented Feb 13, 2026

Merging this PR will not alter performance

Footnotes

Uh oh!

cursor bot left a comment

Choose a reason for hiding this comment

Uh oh!

cursor bot Feb 13, 2026

Choose a reason for hiding this comment

Inverted throughput formula for Bytes measurement scaling

Uh oh!

cursor bot Feb 13, 2026

Choose a reason for hiding this comment

Stale throughput leaks into lookup benchmark results

Uh oh!

github-actions bot commented Feb 13, 2026

Benchmark results

@rust/hash-graph-benches – Integrations

policy_resolution_large

policy_resolution_medium

policy_resolution_none

policy_resolution_small

read_scaling_complete

read_scaling_linkless

representative_read_entity

representative_read_entity_type

representative_read_multiple_entities

scenarios

Uh oh!

indietyp commented Feb 13, 2026

Uh oh!

Reviewers

Assignees

Labels

Milestone

Development

Uh oh!

1 participant

cursor bot commented Feb 13, 2026 •

edited

Loading

vercel bot commented Feb 13, 2026 •

edited

Loading

codecov bot commented Feb 13, 2026 •

edited

Loading